Theoretical framework

Despite the increasing popularity of intergenerational practices, reported developments often lack the use of theory, a strong conceptual framework, and an outcome-driven evidence-base [29, 30]. The developed intervention was underpinned by the components of Contact Theory [29], which supports the benefits of social contact between different groups. Often associated with reductions in prejudice, prior research both within and outside of laboratory settings has focused on its ability to change negative attitudes and reduce the impact or threat of stigma. Comparable outcomes have been reported whether contact is structured or unstructured [31]. However, Allport [29] suggests that we are inherently more sensitive to factors that conform to our negative perceptions and stereotypes, and that casual contact can be superficial and therefore a potential source of increased, not decreased, prejudice. Indeed, it is suggested that “true acquaintance” [29] (p. 264) is needed if stereotypes are to be effectively challenged.

It was postulated that the conceptual principles of Contact Theory could potentially, either directly or indirectly through changes in attitudes and reductions in stereotype threat, be an effective method to target health-related variables in older adults. The work of Allport [29] identified four core components deemed necessary for the successful application of Contact Theory, namely, equal status, cooperation, common goals, and support from social and institutional authorities. A logic model that proposes the potential causal mechanisms by which effects on PA and other health-related outcomes (sedentary behavior and QoL) could be achieved is outlined in Fig 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Conceptual model of technology-driven intergenerational physical activity intervention.

^a BCTs = behavior change techniques; QoL = quality of life; SPA = self-perceptions of aging; VOA = views-on-aging.

https://doi.org/10.1371/journal.pone.0301279.g001

Research design

This feasibility study employed qualitative research methods. A trial period of the proposed intervention of up to four weeks was qualitatively evaluated via post-participation focus groups with the children and older adults. This timeframe was deemed appropriate as four-week trial periods are regularly used in feasibility studies that aim to test the use of technology as a tool in a subsequent intervention trial, as was the case here [32, 33]. Reasons for recruitment limitations were also explored via an additional focus group held with a separately recruited cohort of parents. This data collection addition was made following realization during the recruitment process, that parents were a key link between the older adult and child, and subsequent dyad enrolment.

As this study was designed to test the feasibility of progression to a larger scale trial and is positioned within the developmental stage of the Medical Research Council (MRC) framework for complex interventions [27], no baseline or outcome measures, or exploration of intervention effectiveness was included. Moreover, as this study was conducted prior to the latest MRC guidance version publication, the study design did not include the new recommendation of including an assessment of pre-defined progression criteria. Basic participant demographics (age and gender) and metrics regarding recruitment and retention, were collated to aid the overall analysis. Ethics approval was granted by the Institutional Research Ethics Committee (approval numbers: 2018–103 and 2018-103A). Written, informed consent or assent was obtained from all participants, and the parent/guardian of each participating child. Prior to enrolment, all dyad participants were health screened by the primary researcher (RLK), using a questionnaire based on American College of Sports Medicine (ASCM) guidelines [34] to ensure there were no contradictions to participation. The study methodology, interview schedules and questionnaires were reviewed and discussed by the primary researcher (RLK) with JH, KAM, and AC, who between them have a broad range of experience across older adult and childhood PA, and stereotype-based research disciplines. The primary researcher (RLK), who implemented the intervention trial, conducted the focus groups, and carried out the primary data analysis is a registered healthcare professional with extensive clinical experience. A post-positivistic epistemological position was taken throughout this research. This reality was viewed through a critical realist lens that acknowledges that the ‘real-world’ potentially sits behind subjective and social complexities that shape a truth [35].

Participants

Intergenerational dyad participants were recruited via purposive sampling from a single Primary School in South Wales, United Kingdom. Recruitment packs were sent to all children aged 7–11 years, with additional recruitment posters placed within school grounds and sent out via the schools’ electronic message platform. The older adult (aged ≥ 65 years) could either be a family member of the child (aged 7–11 years), or, an older adult the child had a “familiar” link with. For further inclusion/exclusion criteria see Table 1. In total, six expressions of interest were received, with four dyads consenting to participate. Two dyads were excluded because they were proposing to engage as a parent-child dyad, not a grandparent-child dyad. This, in line with previous qualitative feasibility studies [28], provided an overall sample size of eight individuals, all of whom were white British. Further demographic and participation data for each dyad are presented in Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Intergenerational dyad study inclusion/exclusion criteria.

https://doi.org/10.1371/journal.pone.0301279.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Demographic and participation data of the intervention dyads.

https://doi.org/10.1371/journal.pone.0301279.t002

The additional focus group of parents was purposefully recruited from those who had previously contacted the researcher with queries or for further information during the recruitment process. Four individuals initially agreed to participate; however, one withdrew, resulting in three participants in the final focus group (2 females, 1 male; aged 38–47 years). Each identified at least one older adult within the age range 65–73 years, and, a child aged 8–11 years who would have been eligible to participate in the study.

Intervention

Moving Together is a multi-faceted technology-driven intervention designed to facilitate PA in familial intergenerational dyads. Through engagement with World Walking, dyads combine their daily step counts (recorded via wrist-worn ‘Mi Band 2^©‘ activity trackers) to collaboratively complete pre-developed virtual walk routes. World Walking is an interactive step-count-based, open access, online platform, accessible via a computer, or additional downloaded App. The walk routes, in this instance, were along the coast of Wales in the United Kingdom. Routes are designed to include several target landmark milestones. As steps are added to the walk, the associated map route is updated to outline the percentage completed. When each milestone is reached, information of interest about the location is revealed to the user. For a breakdown of the intervention components, their associated BCTs or ‘active’ elements [36], and modes of delivery see Table 3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Components of the intervention and their associated BCTs and modes of delivery.

https://doi.org/10.1371/journal.pone.0301279.t003

Having been facilitated to form an intergenerational partnership, supplied with the overall collaborative goal of completing the walk route within the trial period, and access to the necessary technology and resources to complete the target goal, dyads are free to set their shared daily target and generate their daily steps using any form of PA or structured exercise recordable by the Mi Band 2^©, (including those generated through activities of daily living). The only stipulation is that both members of the dyad must contribute to the step totals and therefore the completion of the challenge. Activities can be undertaken either together, separately, or a combination of both, and dyads are free to establish their own preferred ways to communicate their achievements with their partner. Table 4 outlines how each parameter of Contact Theory [29] was addressed during intervention development.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Contact Theory [29] intervention components.

https://doi.org/10.1371/journal.pone.0301279.t004

Procedures

Prior to commencing the intervention, the dyads, and, a parent/guardian for each child attended a face-to-face enrolment and induction session with RLK at the Primary School, during mid-March 2019. Consent and health screening were reviewed, and basic demographic data collected. Both members of the intergenerational dyad were supplied with a “Mi Band 2^©” activity tracker, and, assisted to download and set up the associated “Mi Fit App^©” on the smart device of their choice. For the children, this was governed by their parent. For the older adults, access to World Walking was also established; the researcher provided assistance to download and set-up the App as required by each individual, which varied from providing written instructions which were followed independently, to providing practical help with the set-up. Dyads commenced intervention participation immediately following enrolment. No individual exercise prescription occurred, nor was this the intention in any subsequent planned studies. During this feasibility work, participants were only encouraged to engage with and trial the concept of the intervention. The intention of the designed intervention was to provide and stimulate a pragmatic situation replicable in the real-world. This approach, allowing for individual choice, was adopted to allow individual participation levels to be pragmatically explored, when, as per real-world situations, exercise prescription does not routinely occur.

Only the older adult could be provided with access to the World Walking platform due to data protection access restrictions. Therefore, the child separately recorded their step data and liaised with their co-participant for it to be added to their combined totals. The child was supplied with an A3 copy of the map/walk route, to allow them to additionally chart the collaborative progress, as per that shared with them from World Walking, by their partner. To allow for personal preference, each older adult was also supplied with a printed step record sheet, to record their daily steps, instead of referring to the Mi Fit App^©. For copies of the step records, see S1 File. During weeks one to three of the study, each older adult participant and the parent/guardian of each child were briefly contacted via email to discuss any issues or concerns. The study is reported according to the Standards for Reporting Qualitative Research: A synthesis of recommendations (SRQR) checklist [37] (S2 File).

Data collection

In total, three separate researcher-led focus groups were held, one for the child dyad participants, one for the older adult dyad participants, and one for the additionally recruited non-participant parents. The focus groups with the dyad participants were designed to obtain information about participants’ experiences of using and interacting with the fitness tracker and technology platform, and their overall experiences of collaborating and working intergenerationally, to inform future recommendations for implementation within a larger scale trial. Regarding the older adults, the participants’ opinions concerning the impact of the intergenerational collaboration on their views-on-aging, and age stereotypes in general were also explored. The additional focus group undertaken with the recruited non-participant parents discussed participant recruitment issues/limitations, the overall intervention concept, opinions, and potential solutions. For the outline interview schedules for each group, see S3 File.

All focus groups, undertaken following the completion of the intervention trial period at the recruitment Primary School, were conducted by the primary researcher RLK, in a non-direct, neutral manner [38]. The sessions, ranging in duration from 20 to 75 minutes, were audio and video recorded (Phillips Digital Voice Recorder; Sony HandyCam), transcribed verbatim by RLK, and anonymized. Where the focus group involved children, due to the room location, to comply with child safety requirements, a teaching assistant, identified by the school was also in attendance.

Data analysis

Focus group analysis was undertaken drawing on the thematic analysis process outlined by Braun and Clarke [39, 40]. Using this iterative structured approach, the coding process was deductively driven by a pre-defined, study aim-specific, individually constructed thematic framework. Under the headings: acceptability; functionality; useability, and, recruitment and retention, codes were allowed to emerge inductively. Only the semantic meaning of the data, as presented by the participants, was explored [41].

All coding was initially conducted manually by RLK, with subsequent categorization, organization, re-checking, and refinement carried out within Microsoft Word (Office 365^©). Following the initial coding process, a second researcher (JH) blindly cross-matched 10% of the data extracted against the generated codes to ensure consistency in approach and appropriate data coding. From the 10% cross-matched, eight discrepancies were discussed and reviewed back to the original data set until agreement was reached, with the remainder of the initial coding re-reviewed, as required. Codes and sub-themes were challenged and checked back against the original transcripts to ensure data fit, removed and/or rearranged to ensure accurate data representation, and renamed accordingly. These processes were, for transparency, credibility, quality control and rigor [42], completed in collaboration with a ‘critical friend’ (JH). Examples from each stage of the process are provided within the coding audit trail (see S4 File).

Reflexivity for trustworthiness

It is pertinent to note that the research was undertaken in a location where the primary researcher (RLK) was known to some of the participants and the wider Primary School community. Consideration therefore needs to be given to the way in which this might have affected interaction and responses. This particularly related to the non-participant parental group, who all had a priori knowledge of the primary researcher (RLK) and their professional background as a healthcare professional. At all times during data collection impartiality was strived for. As the aim of the research was to establish feasibility, the common codes and sub-themes triangulated from different perspectives across all data sets were afforded equal importance [43].

Acceptability

Functionality

Usability

Recruitment and retention

Intergenerational physical activity: A positive approach?

All participants successfully engaged with the intervention for the whole duration of the trial period indicating acceptability, and, the provision of a potential platform to generate positive behavioral changes and health outcomes. Participants signaled that they enjoyed taking part. Enjoyment, particularly when considering or undertaking more PA has been deemed an important motivational factor for both older adults [45, 46] and children [47]. However, it is pertinent to note that the reasons for impact, and therefore the underpinning mechanisms at work, may be different for dyad members within each targeted age range, in this instance older adults aged ≥ 65 years and children aged 7–11 years.

It was observed that where the older adults appeared to draw direct and explicit associations between daily monitored step counts, self-determined goal progress and success, for children, the effects could be more subtle and implicit. In this situation, the children’s benefits appear to have arisen from their desire to “beat” their co-participant and complete the walk. The application of behavior change models to one population age group, just because they have demonstrated success with another, has already been questioned [2], therefore, the different mechanisms of change and effect, are not surprising. Exploration of the potential impact of different variations of interventions of this nature, to optimize the benefits to all age ranges, could be warranted.

Intergenerational contact provides both children and older adults with the opportunity for generativity, a concept identified as a potentially important driving force in successful partnerships and outcomes [16, 48], and, within this study, a recruitment facilitator. Whilst classically presented as an opportunity to guide and help the next generation [49], through its ability to determine self-worth in later life [50], it appears that in pre-pubescent years, children may also be able to identify and attach comparable benefit to the perceived ability to help others.

Is the stereotype cliché getting old?

Intergenerational contact has been proposed as a way to target the detrimental effects and limit the impact of negative stereotypes of aging across generations [51–53]. The work of Abrams et al. [51] reported that the effects of stereotype threat on older adults aged 59–89 years, were notably suppressed when prior contact with young people had been more positive. Comparison with outgroup members significantly impaired cognitive performance in individuals who experienced less contact, with those who had experienced higher levels of contact relatively unaffected.

Levy et al. [54] found that both implicit and explicit stereotype manipulations led to increased physical function in older adults, however implicit strategies achieved effects that were 30% greater. Intergenerational contact could be viewed as an implicit stereotype manipulation, where, through applying the theoretical components of Contact Theory [29], self-perceptions of ageing, and views-on-ageing in older adults, and attitudes towards ageing in children, are subtly challenged in line with the principles of stereotype embodiment theory [10]. Indeed, although preliminary, the findings of this study indicate that this could be a viable proposition. Alternatively, positive changes could be the result of the construction of an optimum situation and environment, that is context-specific, and again, underpinned by the parameters of Contact Theory [29], leading to reductions in perceived stereotype threat [9]. Ironically, within this feasibility study, negative stereotypes of aging, both self-perceived and views-on-aging, particularly relating to the abilities of older adults to engage with technology and be physically active, were evident. Additional strategies to challenge such stereotypes may need to be an integral part of recruitment processes.

Family: Friend or foe?

Where associations have been made between social support levels, and physical inactivity in older adults [55, 56], the role of the family unit, its structural make-up, and hierarchy within in it, is undoubtedly complex. Important observations were raised regarding the targeted age range inclusion criteria, and the impact this could have on the availability (i.e., due to the generational age gap), and accessibility (i.e., due to the pre-conceived views of the wider family), of the corresponding generations. Allowing ‘younger’ older adults to participate was a suggested solution. Whilst this would potentially change the boundaries of any conclusions that could be drawn, it has been suggested that age stereotypes may actually become less threatening with advancing age [57], and, that salience to stereotype threat in particular, is indeed greater, during the transition into older adulthood [58]. Therefore, intergenerational interventions that target ‘early’ older adulthood could be more effective.

When designing and ultimately endeavoring to implement intergenerational interventions or programs, how members of the wider family will view and engage with the concept also warrants consideration. Strategies to challenge their beliefs and pre-conceived opinions may be essential. Particularly with familial older adults and children, the parental ‘gatekeeper’ who may end up mediating participation, could, albeit unintentionally, considerably help or hinder success [59]. However, the extent of any impact may not always be initially apparent.

How do we solve the problem of recruitment?

Perhaps the biggest challenge facing intergenerational research, particularly where the target populations are older adults and children, is how to effectively recruit sufficient participant numbers. Addressing this issue is of paramount importance, as at present, the evidence-base surrounding this concept and its ability to positively affect health outcomes in older adults is at best, anecdotal. In line with the encountered limitations, prior studies have experienced similar issues. For instance, the iStep project initially aimed to explore the effects of a pedometer-based intergenerational social innovation on obesity levels in older adults through the formation of grandparent/grandchild partnerships. Unsuccessful recruitment led to the formation of pupil/teacher, and pupil/parent partnerships instead [19, 60]. Regardless of the level of potential afforded to an intervention concept or behavioral change strategy, or, how accurately it is constructed, a failure to recruit limits the ability to explore engagement, the magnitude of any observed change, and, the wider transferability of results [61].

There is an apparent need to address the divide between how the research process is presented and subsequently viewed. Targeted, population specific, innovative recruitment strategies need to be devised that evoke interest, demonstrate a positive benefit to burden ratio, and, where necessary, subtly challenge the perceptions and opinions of not only potential participants, but also their wider circle of often influential family and friends [59]. Additionally, specific to interventions utilizing technology, consideration of how terminology is used, and, the connotations that could arise from different interpretations of seemingly standardized wording, for example, ‘activity tracker’ is needed.

Reporting the lessons learnt from recruiting 777 older adult participants, aged ≥ 65 years with a high risk of mobility disability into a 12-month multicenter RCT, Withall et al. [62] adopted a variety of recruitment strategies. The most effective strategy appeared to be mail invites via General Practitioners. Face-to-face recruitment via liaison with and presentations at third-sector organizations (i.e., shelter housing) provided minimal uptake. However, noteworthy is their recommendation that to gain a representative sample, and therefore increase transferability, such relationship building methods may still be essential with minority groups.

What are the key considerations for future work?

For older adults, when designing behavior change interventions, whether intergenerational or not, the findings of this study suggest it may be pertinent to consider whether they allow for flexibility within pre-established routines and individual choice. They also need to account for a potential lack of interest in and engagement with, rigidly imposed structures, and enable ‘activity’ to occur as a by-product of participation in other activities [63]. It is however noted that the effectiveness of self-regulatory techniques with older adults is questionable [2, 64]. Within this study, attainable achievement was also perceived as being important, where the distance between some of the milestones within World Walking was deemed to be too far, motivation waned.

Another key ‘gripe’, that could have been a contributory factor to the issue that arose whereby some parents were wearing their children’s watches, was the inability to record steps for other sports and activities undertaken. Consideration needs to be given to how this issue can be addressed. One possibility could be the provision of activity-to-step conversion charts, thus allowing the accumulation and addition of equivalent step data. Less despondence from the children could remove the need for parents to feel compelled to help. Some purposefully constructed, study specific, web-platforms have utilized more sophisticated built-in step calculators [65], however, within real-world research, options are restricted.

Strengths and limitations

Within this study, 75% of focus group participants and 100% of parental responses were female. Positively, each focus group did have one male representative. This is important as it has been noted that gender disparities within such research are common [15], and, that males and females could have different outcome responses to intergenerational interventions and activities [53]. Criticism may also be drawn towards the small convenience sample obtained from only one local school and hence the limited scope for transferability of the results. However, given the identified recruitment complexities, and failure of other studies to recruit any of their targeted sample [19, 60], this work still provides some valuable insight. Additionally, the strategies employed within the interview schedules with the children appeared to be effective. Given the young age of the participants, and their developing linguistic ability [66], the use of monosyllabic responses was minimal, and the volume of data obtained deemed sufficient. It is, however, pertinent to note, that for some of the sub-themes (i.e., engagement) the supporting data is more heavily reflective of adult, rather than child, perspectives. In line with the study aims, the interview schedule questions focused on the novel technology aspects, rather than the context-specific behavior change strategies and their implementation. The inclusion of questions on these parameters could have broadened the depth of knowledge generated. Finally, the dyads only experienced a relatively short trial period. It is therefore difficult to truly understand the implications of any novelty factor or longer-term adherence issues. Whilst the developed intervention was underpinned by the components of Contact Theory [29], future large-scale interventions are required to ascertain how Contact Theory relates to the findings.

It is pertinent to note that the primary researcher was a parent at the participating school, and, the only focus group moderator. Whilst professional and ethical boundaries were always observed, an impact on recruitment, and, the subsequent results, albeit positive or negative, cannot be ruled out. To increase rigor and trustworthiness, within the focus groups, outline interview question schedules were specified a priori, and, within the data analysis, identified codes and themes were reflected on and challenged by a critical friend to ensure that the personality, experiences, and, beliefs of the researcher, and, goals of the research did not influence the analysis and reporting.